Hi Armin,
Thanks for spending hours of investigation.
I think it is absolutely correct, that OJB uses only
one PersistenceBroker and as you described it seems
that it is a problem of JBoss.
Again thanks a lot for your detailed description.
regards,
André
-Original Message-
From:
I found out that the problem occurs if the JavaServer Faces framework
serializes an object and reloads it afterwards.
Is it forbidden to serialize and recreate OJB-created objects with
proxies in it? Do I have to mark the lists created by OJB as transient?
But if they are transient, on
Hi Martin,
I try to reproduce your problem without success. Could you please post a
test case or detail describe (some pseudo code) what your are doing?
By the way, do you use anonymous keys in collection objects?
regards,
Armin
Martin Marinschek wrote:
I found out that the problem occurs if the
Marcus Young wrote:
Each class is mapped to a separate table.
I persuaded, that it's much easier to map the whole hierarchy to one
table and provide ojbConcreteClass..
David
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For
Hi Marcus,
did you try to declare class-descriptor's for B and C (including all
fields and references) and define an extent for class A:
class-descriptor class=A
extent-class class-ref=B /
extent-class class-ref=C /
/class-descriptor
regards,
Armin
Marcus Young wrote:
Hi,
I have just been
David,
thanks for your comments. The main problem that I
see with this approach is that each of the classes
inheriting from the abstract class have very
different attributes - so there will be a lot of
empty fields in the table - I am not an expert,
however I would assume that this is not
Hi Armin,
yes - I tried defining an extent and the appropriate
class-descriptor's - I also experimented with
super. I assume that I am correct in assuming
that it is possible to build such a structure with
an abstract class?
regards,
Marcus
Hi Marcus,
did you try to declare
Martin Marinschek wrote:
I am not using anonymous keys at all, afaik every object in my system
has its id and exposes it also in the source-code.
ok, similar to my local test
Here is my class-descriptor for the class being serialized and loaded
back; I realized that I am using the
Armin Waibel wrote:
Martin Marinschek wrote:
I am not using anonymous keys at all, afaik every object in my system
has its id and exposes it also in the source-code.
ok, similar to my local test
Here is my class-descriptor for the class being serialized and loaded
back; I realized that I am
good and bad table design is disputable, but in fact; if you use Oracle
Designer for example, they use exactly this approach for all
generalization-relationships by default.
-Martin
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
Martin,
on thinking about it - a single table may be the
best approach. One of my objectives is to be able
to build a collection of different objects (all from
the same base class) - a sort of recursive
collection. It seems that ojbConcreteClass would
provide support for this - I'm not
I got rid of the factory and the extended class, and the stuff works now.
Thanks for your help,
-Martin
Martin Marinschek wrote:
Armin Waibel wrote:
Martin Marinschek wrote:
I am not using anonymous keys at all, afaik every object in my
system has its id and exposes it also in the source-code.
[EMAIL PROTECTED] wrote:
Hi Armin,
yes - I tried defining an extent and the appropriate
class-descriptor's - I also experimented with
super. I assume that I am correct in assuming
that it is possible to build such a structure with
an abstract class?
yep, I wrote a test case (using extent and
Try using the sequence manager that uses a stored procedure 'oracle-style'
(I think it's called SequenceManagerStoredProcedureImpl - what's what we are
using with SQL Server 2000 (we have 4 app servers and other misc vms tied
into it). I'd also recommend the jgroups/oscache distributed cache.
14 matches
Mail list logo